The Fish Pathogen Vibrio Ordalii Under Iron Deprivation Produces the Siderophore Piscibactin

microorganisms

Article The Fish Pathogen Vibrio ordalii Under Iron Deprivation Produces the Siderophore Piscibactin

Pamela Ruiz 1,2, Miguel Balado 3, Juan Carlos Fuentes-Monteverde 4 , Alicia E. Toranzo 3, Jaime Rodríguez 4 , Carlos Jiménez 4 , Ruben Avendaño-Herrera 1,2,* and Manuel L. Lemos 3,* 1 Laboratorio de Patología de Organismos Acuáticos y Biotecnología Acuícola, Facultad de Ciencias de la Vida, Universidad Andrés Bello, 2531015 Viña del Mar, Chile 2 Interdisciplinary Center for Aquaculture Research (INCAR), 2531015 Viña del Mar, Chile 3 Departamento de Microbiología y Parasitología, CIBUS-Facultad de Biología and Instituto de Acuicultura, Universidade de Santiago de Compostela, 15782 Santiago de Compostela, Spain 4 Centro de Investigacións Cientíﬁcas Avanzadas (CICA), Departamento de Química, Facultade de Ciencias, Universidade da Coruña, 15071 A Coruña, Spain * Correspondence: [email protected] (R.A.-H.); [email protected] (M.L.L.)

Received: 31 July 2019; Accepted: 31 August 2019; Published: 3 September 2019

Abstract: Vibrio ordalii is the causative agent of vibriosis, mainly in salmonid fishes, and its virulence mechanisms are still not completely understood. In previous works we demonstrated that V. ordalii possess several iron uptake mechanisms based on heme utilization and siderophore production. The aim of the present work was to confirm the production and utilization of piscibactin as a siderophore by V. ordalii. Using genetic analysis, identification by peptide mass fingerprinting (PMF) of iron-regulated membrane proteins and chemical identification by LC-HRMS, we were able to clearly demonstrate that V. ordalii produces piscibactin under iron limitation. The synthesis and transport of this siderophore is encoded by a chromosomal gene cluster homologous to another one described in V. anguillarum, which also encodes the synthesis of piscibactin. Using β-galactosidase assays we were able to show that two potential promoters regulated by iron control the transcription of this gene cluster in V. ordalii. Moreover, biosynthetic and transport proteins corresponding to piscibactin synthesis and uptake could be identified in membrane fractions of V. ordalii cells grown under iron limitation. The synthesis of piscibactin was previously reported in other fish pathogens like Photobacterium damselae subsp. piscicida and V. anguillarum, which highlights the importance of this siderophore as a key virulence factor in Vibrionaceae bacteria infecting poikilothermic animals.

Keywords: Vibrio ordalii; ﬁsh pathogens; iron uptake; siderophores; piscibactin; vanchrobactin

1. Introduction Vibrio ordalii is a γ-proteobacterium which causes vibriosis, a hemorrhagic septicemia, in several species of aquacultured fish, mainly salmonids [1]. Although vibriosis outbreaks due to V. ordalii have been reported around the globe, in the last 15 years they reached an important impact in Chile, where they cause significant economic losses in salmonids aquaculture [2,3]. Besides its genetic similarity to V. anguillarum [4,5], another important fish pathogen with worldwide distribution, many aspects of the virulence mechanisms of V. ordalii still remain unknown. While its pathogenicity is not correlated to erythrocytes hemagglutination capacity or biofilm formation in Atlantic salmon (Salmo salar), the hydrophobic properties of V. ordalii cells could play a role in virulence. Moreover, V. ordalii can evade the host immune system and can survive within Atlantic salmon mucus, which likely facilitates colonization [3,6]. However, many aspects of its ability to colonize and multiply within the fish hosts remain unclear.

Microorganisms 2019, 7, 313; doi:10.3390/microorganisms7090313 www.mdpi.com/journal/microorganisms Microorganisms 2019, 7, 313 2 of 16

For most bacteria iron uptake ability during the naturally iron-limited conditions of an infection is a key virulence factor essential for multiplication within the host [7–9]. Besides the importance of iron for the cell metabolism, this element is an important signal that regulates expression of many other metabolic and virulence functions in bacterial cells [10]. This regulation is usually mediated by the transcriptional regulator Fur which needs Fe2+ as cofactor to bind to the promoter region of genes controlled by iron levels and prevent the binding of RNA polymerase to DNA [11]. The main mechanisms described in Gram-negative bacteria to get iron from the cell surroundings are the direct use of heme groups as a source of iron [12] and the synthesis of siderophores, which can efficiently sequester the iron bound by transferrins and other iron-holding proteins within the host [9,13,14]. The ferri-siderophore is then internalized through specific TonB-dependent outer membrane protein receptors that are energized through the TonB system [15–17]. Bacterial fish pathogens are not an exception for iron requirements and several mechanisms of iron uptake, including the use of heme and the synthesis of siderophores, have been reported in many of these bacteria [18–25]. We have previously demonstrated that V. ordalii can also use heme and hemoglobin as iron sources and that it has the ability to produce siderophores [26]. However, despite the clear relationship between V. ordalii iron uptake ability and pathogenicity, the precise nature of the iron assimilation mechanisms remains unclear. In this previous work, from genetic and genomic analysis, the results of cross-feeding assays, and from some other data in the literature [4], we suggested that V. ordalii could likely produce piscibactin as a siderophore. Piscibactin was isolated and characterized from the fish pathogen Photobacterium damselae subsp. piscicida [23]. In this bacterium piscibactin synthesis is encoded in a pathogenicity island harbored in the pPHDP70 virulence plasmid [27]. Recent in silico genomic studies in the Vibrionaceae family showed that the gene cluster encoding piscibactin synthesis and transport is really widespread in many species of Vibrio and Photobacterium [28]. In fact, we have recently demonstrated that some strains of V. anguillarum, a bacterium closely related to V. ordalii, produces piscibactin in a temperature-dependent fashion, being preferentially expressed at low temperatures. In these conditions piscibactin synthesis is a key virulence factor for V. anguillarum [29]. In the present work, we have characterized the gene cluster encoding the biosynthesis and transport of piscibactin and demonstrated, by genetic, proteomic and chemical analysis, that piscibactin is indeed produced as siderophore by V. ordalii.

2. Materials and Methods

2.1. Bacterial Strains and Growth Conditions Three V. ordalii strains were used: The type strain ATCC 33509T and two strains, Vo-LM-13 and Vo-LM-18, previously isolated from vibriosis outbreaks in Atlantic salmon cultured in Chile [3,6]. All were conﬁrmed as V. ordalii according to the PCR protocol previously described [30]. All strains were routinely cultivated on Trypticase Soy Agar or Trypticase Soy Broth supplemented with 1% (w/v) NaCl (TSA-1 and TSB-1, respectively). For some experiments the CM9 minimal medium was also used [31]. Stock cultures were kept frozen at 80 C in Criobilles tubes (AES Laboratories, Combourg, France) or − ◦ in TSB-1 with 15% (v/v) glycerol.

2.2. RNA Extraction and RT-PCR To analyze the transcriptional regulation of the gene cluster involved in the biosynthesis and transport of the siderophore piscibactin, a RT-PCR was performed with the primers listed in Table1. For this assay, V. ordalii Vo-LM-18 was grown in iron-limited (TSB-1 plus 2,20-dipyridyl), iron-excess (TSB-1 plus FeCl3 10 µM) and standard conditions (TSB-1). Total RNA was prepared from cultures after 48 h post-incubation using TRIzol® reagent (Ambion-ThermoFisher, Waltham, MS, USA) according to the manufacturer’s instructions. Each RNA sample was subjected to treatment with DNase I RNase free. To obtain the cDNA, 5 µg total RNA and reverse transcriptase enzyme M-MLV (Invitrogen-ThermoFisher, Waltham, MS, USA) was used following the manufacturer’s instructions for each reverse transcription Microorganisms 2019, 7, 313 3 of 16 reaction. The PCR reaction was prepared with the cDNA, 1 U of BioTaq DNA polymerase (Bioline, Memphis, TN, USA), 200 µM of each dNTP and 2 mM MgCl2, ﬁnal concentration. Depending on the melting temperature (Tm) of each pair of primers, annealing temperatures ranged from 55 to 60 C. Times of elongation were selected based on the expected size of ampliﬁcation (1 min kb 1). In all ◦ · − cases, the same reaction mixture, but without reverse transcriptase, was used as negative control, and chromosomal DNA of the Vo-LM-18 strain was used as positive control.

Table 1. Primers used in this work.

Primers Sequence (50-30) * Ampliﬁed Fragment (bp) Ampliﬁcation of potential promoters P1 Promoter 1_F GCGTCTAGACACTTTGCCACCCACCATTA 879 Promoter 1_R GCGGGATCCACGAATCGTCGTGTTGGCAT P2 Promoter 2_F GCGTCTAGACCGCTTAGAGAAACCAACGT 1165 Promoter 2_R GCGGGATCCACGTTTCGGTAAGCGTATGG Transcriptional regulation of irp gene cluster RT TTTGGAGATGAGTGCGACAC PCR1 ARC1ordalii_F GATATGCGCTTTGACTGCCA 196 ARC1ordalii_R CTGTGAGACGGCATACAAGC PCR2 FrpA_ordalii_F CGGTGGTAATGCTCAAGGTG 204 FrpA_ordalii_R TGGCTCGGTAGGTGTTCAAT PCR3 Irp2_ordalii_F AGCAGGCAACAAAGAGTGAG 413 Irp1_ordalii_R GGGCGAATAACCAAACAAGC * Recognition sequences for restriction enzymes are underlined.

2.3. Construction of lacZ Transcriptional Fusions and β-Galactosidase Assays The presence of potential gene promoters within the piscibactin gene cluster of V. ordalii was performed using BPROM tool [32]. Putative Fur boxes were detected by an in silico search of the GATAAT hexamer [33]. DNA fragments corresponding to V. ordalii frpA and araC1 promoter regions (P1 and P2, respectively) were amplified by PCR using primers specified in Table1. The amplified fragments included the region upstream of the start codon and the first nucleotides (ca. 50 bp) of frpA or araC1 coding sequences. These putative promoter regions were fused to a promoterless lacZ gene and inserted into the low-copy-number reporter plasmid pHRP309 [34]. The resulting transcriptional fusion constructs, P1::lacZ and P2::lacZ, were mobilized from Escherichia coli β3914 into V. ordalii Vo-LM-18 by conjugation. Transformed ex-conjugants were selected on the basis on their resistance to gentamicin (pHRP309 marker). As a negative control, V. ordalii Vo-LM-18 with an empty pHRP309 was used. To determine whether potential promoters were regulated by iron, a total of four growth conditions were tested for each one of the transcriptional fusions: Cells grown in CM9, cells grown under iron excess (CM9 plus FeCl3 20 µM) and two iron limiting conditions, CM9 plus 2,20-dipyridyl 25 mM and CM9 plus 2,20-dipyridyl 80 µM. All cultures were carried out with agitation at 100 rpm at 18 ◦C until an OD600~0.1 to record the β-galactosidase activity. The transcriptional activity was determined by measuring the β-galactosidase activity of fusions P1::lacZ and P2::lacZ following the method described by Miller [35]. Volumes of 0.1 and 0.5 mL, respectively, were used. Both were brought to a final volume of 1 mL with buffer Z (Na2HPO4 2H2O Microorganisms 2019, 7, 313 4 of 16

60 mM; NaH PO H O 40 mM; KCl 10 mM; MgSO 7H O 1 mM and β-mercaptoethanol 50 mM; 2 4· 2 4 2 pH 7.0). To this mixture 20 µL of chloroform and 10 µL of a solution of 0.1% SDS were added and the ﬁnal solution was incubated at 37 ◦C for 5 min. The reaction was initiated by adding 0.2 mL of ortho-nitrophenyl-β-galactoside (ONPG; 4 mg mL 1 in Z buﬀer). The reaction was stopped with · − 0.5 mL of 1 M Na2CO3 when a color change to yellow was generated. Finally, A420 was measured in a UV-VIS spectrophotometer (Hitachi U2000, Tokyo, Japan).

2.4. Analysis of Outer Membrane Proteins (OMP) Profile of V. ordalii OMPs were obtained from V. ordalii strains ATCC 33509T, Vo-LM-18, and Vo-LM-13 grown under iron excess (TSB-1) and iron limitation (TSB-1 + 2,20-dipyridyl, using a concentration half of the specific MIC for each strain). Each strain was cultured in 500 mL of TSB-1 or TSB-1 + 2,20-dipyridyl at 18 ◦C for 48 h. After incubation, the media were centrifuged at 10,000 g for 10 min at 4 C. The cell pellets × ◦ were resuspended in 3 mL of a solution containing 10 mM Tris-HCl (pH 8.0), 0.3% NaCl and 1% of a protease inhibitor cocktail (Sigma-Aldrich, St. Louis, MO, USA). The suspension was then sonicated three times with a Branson 250 Sonifier (60 W pulses for 30 s, 30 s intervals in ice). After 1–2 min of centrifugation to eliminate cell debris, supernatants were centrifuged at 17,000 g for 60 min at 4 C. × ◦ The pellets obtained contained total cell membranes. Outer membrane fractions were obtained as previously described [36,37]. Briefly, the total membrane pellets were resuspended in a solution containing 20 mM Tris-HCl (pH 8.0), 3% (w/v) sodium lauryl sarcosinate (Sigma-Aldrich, St. Louis, MO, USA) and 1% protease inhibitor cocktail (Sigma-Aldrich, St. Louis, MO, USA). The suspension was incubated at room temperature for 20 min to dissolve the inner membrane. Outer membranes were pelleted by 100,000 g ultracentrifugation × for 60 min at 4 ◦C and washed twice with distilled water. Protein concentration was determined using the BCA Assay Kit (Thermo Scientific, Waltham, MS, USA), and samples were kept at 20 C until use. − ◦ Iron-regulated OMP (IROMP) profiles were compared for each V. ordalii strain between cells grown with or without iron limitation. Each extract (20 µg) was mixed with the SDS-PAGE sample buffer, heated at 95 ◦C for 5 min, and separated by SDS-PAGE with 7.5% (w/v) acrylamide in the resolving gel. Electrophoresis was performed in a Mini-PROTEAN 3 Cell (Bio-Rad, Portland, ME, USA) at 120 V for 120 min. Protein bands were stained with 0.05% Coomassie blue R (Sigma-Aldrich, St. Louis, MO, USA) for at least 1 h and destained for 2 h in 10% methanol and 10% acetic acid. The relative mobility of each protein was determined by comparison with standard protein markers (Precision Plus Protein Standards, Bio-Rad). Digital images were collected using a G:BOX Chemi XT4 Fluorescent and Chemiluminescent Imaging System (Syngene, Frederick, MD, USA) with GeneSys automatic control software and GeneTools analysis software (Syngene, Frederick, MD, USA). Three independent separations, from two different cultures, were performed for each strain and growth condition. Further analyses were done only to protein bands induced or increased in intensity under iron-limited conditions.

2.5. Protein Identification by Peptide Mass Fingerprinting (PMF) Candidate iron-regulated bands from SDS-PAGE gels were identified by PMF, using MALDI-TOF (Matrix-Assisted Laser Desorption/Ionization Time-of-Flight) Mass Spectrometry analysis, as previously described [38]. In brief, protein bands of interest were manually excised and subjected to in-gel digestion with trypsin using the In-Gel DigestZp Kit (Millipore ES, Madrid, Spain), following the manufacturer’s protocol, to extract proteins prior to mass spectrometry analysis. Before digestion, the samples were reduced with dithiothreitol and alkylated with iodoacetamide. Proteins were identified by PMF with an Ultraflex III TOF/TOF (Bruker ES, Madrid, Spain). For negative identifications, due to mixed proteins in a single band, a liquid chromatography ion-trap mass-spectrometer system with an amaZon speed ETD (Bruker ES, Madrid, Spain) was used. The SwissProt and NCBInr protein databases were screened with Mascot v2.3 (Matrix Science). The identified peptides were then subjected to a BLASTP analysis using the NCBI (National Center for Biotechnology Information) database, to search for homologues. Microorganisms 2019, 7, 313 5 of 16

2.6. Bioinformatics Tools The DNA and protein sequences were analyzed using the NCBI databases through the BLAST algorithms. The protein families database (Pfam 31.0) of EMBL-EBI (European Bioinformatics Institute) was used to predict the protein domain organization [39]. The functional promoters were identiﬁed using the online database BPROM. The organization of putative domains in biosynthetic proteins were detected using the PKS/NRPS database (http://nrps.igs.umaryland.edu/).

2.7. Detection of Siderophore Piscibactin Piscibactin was detected as previously described [23] with slight modifications as follows: 1 L of cell-free culture broth of strain Vo-LM-18 was concentrated under vacuum (39 ◦C) until 300 mL. Then, 150 mL were transferred to a flat-bottom flask provided with a magnetic stir bar and 750 µL of a solution of GaBr3 in H2O (12 mg/mL) were added dropwise over 5 min and gently stirred for another 10 min. This solution was stored at 4 ◦C during 24 h. An aliquot of the solution containing piscibactin-Ga(III) complex (75 mL) was submitted to Solid Phase Extraction (SPE) through an OASIS® (Waters, Cerdanyola del Vallès, Spain) cartridge (35 cm3, 6 g) using an extraction vacuum manifold (0.2 bar) and eluted with 30 mL of the following mixtures of H2O and CH3CN: 1:0; 1:3; 1:1; 0:1. Fractions were dried out under reduced pressure and subjected to LC-HRMS analysis using an Atlantis dC18 (100 mm 4.6 mm, 5µm) column (Waters) at a flow rate of 1 mL/min. Separation, with a sample × injection volume of 20 µL, was achieved by a 35 min gradient from 10% to 100% of CH3CN in H2O, then a 5 min isocratic step of 100% CH3CN. LC-ESI(+)-HRMS analysis of the fraction eluted with the mixture H2O/CH3CN 1:1, named as L3, showed a peak at 12.06 min that displayed the characteristic isotopic cluster of piscibactin-Ga(III) complex at m/z 518.9928/521.9913. Results are reported following the identification requirements for MS techniques SANTE/11945/2015. Since it was possible detect both ions at significant intensity, the difference between the calculated and the detected exact mass of piscibactin-Ga(III) complex in ppm (∆m/z) and the isotopic ratio abundance error (δ RIA) of M + 1/M could be obtained using the Formulas (1) and (2). The quality of the spectral information was achieved by narrowing the detection m/z range around the compound of interest of 350–600 dalton measured in a LTQ-Orbitrap, which is in agreement with the small values δ RIA found. Formula (1)—SI: Mass accuracy:

m m measured m theoretical ∆ = − 106ppm . (1) z m theoretical ×

Formula (2)—SI: Isotopic ion abundance ratio error (δ RIA):

RIAexp RIAtheo δ RIA(%) = 100 − . (2) × RIAtheo

Presence of the siderophore vanchrobactin in the same cell-free culture supernatants was also detected using the methodology previously described [19,29].

2.8. Statistical Analysis Data from all assays were statistically analyzed using analysis of variance (ANOVA). Signiﬁcant diﬀerences were established as p < 0.05.

3. Results

3.1. Characterization of the V. ordalii Gene Cluster Encoding a Piscibactin-Like Siderophore An in silico analysis of the genome of V. ordalii ATCC 33509 shows the presence of a gene cluster homologous to the piscibactin cluster (irpang) described in the chromosome II of V. anguillarum RV22 [29]. Microorganisms 2019, 7, 313 6 of 16

Both clusters show a high degree of synteny and a similarity between 96% and 98% at the amino acid level (Figure1). This genomic island includes the 11 genes ( irp genes) previously identiﬁed as Microorganismspart of the plasmid 2019, 7, x FOR encoding PEER REVIEW piscibactin in P. damselae subsp. piscicida [27] (Figure1). An in 6silico of 15 search in GenBank, and previously published works [28], show the presence of homologous gene search in GenBank, and previously published works [28], show the presence of homologous gene clusters in several members of the Vibrionaceae family, such as V. cholerae, V. mimicus, V. coralliilyticus, clusters in several members of the Vibrionaceae family, such as V. cholerae, V. mimicus, V. V. anguillarum or Photobacterium profundum. It is noteworthy that this gene cluster exhibits about a 40% coralliilyticus, V. anguillarum or Photobacterium profundum. It is noteworthy that this gene cluster similarity with the genes of the HPI pathogenicity island (encoding the synthesis of the siderophore exhibits about a 40% similarity with the genes of the HPI pathogenicity island (encoding the yersiniabactin) of Yersinia spp. and it was reported as a key virulence factor for P. damselae subsp. synthesis of the siderophore yersiniabactin) of Yersinia spp. and it was reported as a key virulence piscicida [27,40]. factor for P. damselae subsp. piscicida [27,40].

T Figure 1.1. ComparativeComparative analysis analysis of theof Vibriothe Vibrio ordalii ordaliiATCC 33509ATCCT irp33509gene clusterirp gene with cluster the homologous with the homologouschromosomal chromosomal region of V. anguillarum region of RV22V. anguillarum and with theRV22 homologous and with sequencethe homologous from plasmid sequence pPHDP70 from plasmidfrom P. damselae pPHDP70subsp. frompiscicida P. damselae. Biosynthetic subsp. piscicida and regulatory. Biosynthetic genes and are regulatory depicted in genes blue andare depicted the gene inencoding blue and the the outer gene membrane encoding the receptor outer (FrpA)membrane in green. receptor Other (FrpA) genes in andgreen. short Other ORFs genes are and shown short in ORFsclear greenare shown and orange in clear colors. green Grey and blocksorange indicate colors. Gr percentagesey blocks ofindicate similarity percentages in the proteins of similarity sequence. in theThe proteins GenBank sequence. accession numbersThe GenBank and the accession nucleotide numb positionsers and interval the nucleotide are indicated positions below interval the name are of indicatedeach species. below the name of each species.

Piscibactin is synthetized by NRPS-type (non-ribosomal peptid peptidee synthetases) enzymes encoded by irp1irp1 andandirp2 irp2genes genes [23 [23].]. The The bioinformatic bioinformatic analysis analysis of theof the resulting resulting proteins proteins Irp1 andIrp1 Irp2and ofIrp2V. ordaliiof V. ordaliishowed showed that the that catalytic the catalytic domains domains present present in these in enzymes these enzymes are almost are identical, almost identical, with a similarity with a similarityin the amino in the acid amino sequence acid of sequence 99%, to theirof 99%, counterparts to their counterparts encoded by encodedirpang cluster by irp ofangV. cluster anguillarum of V. anguillarumRV22 [29] (Figure RV222 [29]). Thus, (Figure the resulting2). Thus, siderophore the resulting encoded siderophore by the encodedirp cluster by of theV. ordaliiirp clustershould of beV. ordaliialso piscibactin. should be also piscibactin. Like in V. anguillarum, the irp cluster genes of V. ordalii encode most functions needed for piscibactin synthesis and utilization, although an entD homologue is absent in this gene cluster when compared to the P. damselae subsp. piscicida irp cluster (Figure1). The entD gene encodes a 4’-phosphopantetheinyl transferase that is required to activate the peptide synthesis domains of non-ribosomal peptide synthetases (NRPS) [41] and it is essential for piscibactin biosynthesis [23]. However, a homologue of this gene is present in the genome of V. ordalii as part of the vab gene cluster, encoding the siderophore vanchrobactin [4,26]. This entD homologue could provide in trans the function of a 4’-phosphopantetheinyl transferase necessary for piscibactin biosynthesis in V. ordalii. We have previously shown that although vanchrobactin could be synthetized by V. ordalii, it cannot be used as siderophore since the ABC transporters necessary for ferric vanchrobactin internalization are not present in the genome of V. ordalii [26].

Figure 2. Representation of the catalytic domains predicted in Irp1 and Irp2 enzymes of Photobacterium damselae subsp. piscicida, V. ordalii ATCC 33509T and V. anguillarum RV22. Analysis of domains was performed using the PKS/NRPS database (http://nrps.igs.umaryland.edu/). Abbreviations: AT, acyltransferase; Cy, cyclization; KS, ketoacil synthase; KR, ketoreductase; PP, peptidyl-carrier protein; TE, thioesterase. Dotted boxes highlight the main differences.

Microorganisms 2019, 7, x FOR PEER REVIEW 6 of 15 search in GenBank, and previously published works [28], show the presence of homologous gene clusters in several members of the Vibrionaceae family, such as V. cholerae, V. mimicus, V. coralliilyticus, V. anguillarum or Photobacterium profundum. It is noteworthy that this gene cluster exhibits about a 40% similarity with the genes of the HPI pathogenicity island (encoding the synthesis of the siderophore yersiniabactin) of Yersinia spp. and it was reported as a key virulence factor for P. damselae subsp. piscicida [27,40].

Figure 1. Comparative analysis of the Vibrio ordalii ATCC 33509T irp gene cluster with the homologous chromosomal region of V. anguillarum RV22 and with the homologous sequence from plasmid pPHDP70 from P. damselae subsp. piscicida. Biosynthetic and regulatory genes are depicted in blue and the gene encoding the outer membrane receptor (FrpA) in green. Other genes and short ORFs are shown in clear green and orange colors. Grey blocks indicate percentages of similarity in the proteins sequence. The GenBank accession numbers and the nucleotide positions interval are indicated below the name of each species.

Piscibactin is synthetized by NRPS-type (non-ribosomal peptide synthetases) enzymes encoded by irp1 and irp2 genes [23]. The bioinformatic analysis of the resulting proteins Irp1 and Irp2 of V. ordalii showed that the catalytic domains present in these enzymes are almost identical, with a similarity in the amino acid sequence of 99%, to their counterparts encoded by irpang cluster of V. Microorganismsanguillarum RV222019, 7 ,[29] 313 (Figure 2). Thus, the resulting siderophore encoded by the irp cluster 7of of V. 16 ordalii should be also piscibactin.

Figure 2.2.Representation Representation of theof catalyticthe catalytic domains domains predicted predicted in Irp1 and in Irp2Irp1 enzymes and Irp2 of Photobacterium enzymes of damselaePhotobacteriumsubsp. damselaepiscicida subsp, V. ordalii. piscicidaATCC, V. ordalii 33509 ATCCT and 33509V. anguillarumT and V. anguillarumRV22. Analysis RV22. ofAnalysis domains of wasdomains performed was usingperformed the PKS using/NRPS databasethe PKS/NRPS (http://nrps.igs.umaryland.edu database (http://nrps.igs.umaryland.edu/)./). Abbreviations: AT, acyltransferase;Abbreviations: AT, Cy, cyclization;acyltransferase; KS, ketoacil Cy, cyclization; synthase; KS, KR, ketoacil ketoreductase; synthase; PP, peptidyl-carrierKR, ketoreductase; protein; PP, TE,peptidyl-carrier thioesterase. protein; Dotted boxesTE, thioesterase. highlight the Dotted main boxes diﬀerences. highlight the main differences.

3.2. Transcriptional Analysis and Iron Regulation of the Irp Gene Cluster of V. ordalii To test if irp genes of V. ordalii were expressed, several RT-PCR (reverse-transcriptase PCR) reactions were performed. The results showed that the irp gene cluster is transcribed as a polycistronic mRNA that includes araC1, araC2, frpA, irp1-5, irp8 and irp9 genes (Figure3). Therefore, all genes putatively encoding the synthesis, regulation and transport of piscibactin could be co-transcribed from the promoter P2 located upstream of araC1 (Figure3a). An identical result was found for the piscibatin irpang cluster described in V. anguillarum RV22 [29]. This promoter contains a putative Fur box that would indicate that its activity is regulated by the transcriptional regulator Fur in an iron-dependent fashion [33]. An additional promoter P1, also containing a putative Fur box, was located upstream of frpA (Figure3a). The frpA gene would encode the presumptive ferri-piscibactin outer membrane receptor while araC1 would encode a putative AraC-type transcriptional regulator. Thereby, even though irp genes can be transcribed mainly from the promoter upstream of araC1, the existence of additional active promoters cannot be ruled out. In order to analyze the expression levels of the irp putative promoters P1 and P2, DNA fragments of ca. 700 nucleotides upstream of frpA and araC1 genes (Figure3a) were cloned into the plasmid pHRP309 upstream of a promoterless lacZ gene. Resulting plasmids were mobilized into V. ordalii Vo-LM-18 and the transcription levels of lacZ were measured by determining β-galactosidase activity under diﬀerent conditions of iron availability (Figure4). The use of the P frpA (P1) and ParaC1 (P2) presumptive promoters produced signiﬁcant β-galactosidase activity when cells were cultured under a strong iron limitation (CM9 medium plus 2,20-dipyridyl 80 µM). Under iron excess conditions (CM9 or CM9 plus FeCl3 25 µM) the β-galactosidase activity of the P2 promoter was 75% of the P1 promoter (Figure4), suggesting a higher basal activity for this promoter. However, under strong iron limitation, the P2 promoter seems to be 10% more active than P1, suggesting a tighter control by iron levels. These results demonstrate that the two promoter sequences could serve as transcriptional starts of the whole irp operon, and that both of them are strongly regulated by iron levels with slight variations between them. Microorganisms 2019, 7, x FOR PEER REVIEW 7 of 15

Like in V. anguillarum, the irp cluster genes of V. ordalii encode most functions needed for piscibactin synthesis and utilization, although an entD homologue is absent in this gene cluster when compared to the P. damselae subsp. piscicida irp cluster (Figure 1). The entD gene encodes a 4’-phosphopantetheinyl transferase that is required to activate the peptide synthesis domains of non-ribosomal peptide synthetases (NRPS) [41] and it is essential for piscibactin biosynthesis [23]. However, a homologue of this gene is present in the genome of V. ordalii as part of the vab gene cluster, encoding the siderophore vanchrobactin [4,26]. This entD homologue could provide in trans the function of a 4’-phosphopantetheinyl transferase necessary for piscibactin biosynthesis in V. ordalii. We have previously shown that although vanchrobactin could be synthetized by V. ordalii, it cannot be used as siderophore since the ABC transporters necessary for ferric vanchrobactin internalization are not present in the genome of V. ordalii [26].

(a)

(b)

Microorganisms 2019, 7, x FOR PEER REVIEW 8 of 15

reactions lacking reverse transcriptase. Positive controls (+) are PCR reactions using chromosomal DNA as template, +Fe: RT-PCR performed with cells grown under iron excess (TSB-1 + FeCl3 20 µM); -Fe: RT-PCR performed with cells grown under iron limitation (TSB-1 + 2,2’-dipyridyl 60 µM). FigureFigure 3. Transcriptional3. Transcriptional organization organization of of the the gene gene cluster cluster putatively putatively encodingencoding biosynthesisbiosynthesis and In order to analyze the expression levels of the irp putative promoters P1 and P2, DNA transporttransport of siderophore of siderophore piscibactin piscibactin in V.in ordaliiV ordalii.(.a ()a The) The predicted predicted gene gene functions functions are:are: biosynthesis,biosynthesis, fragments of ca. 700 nucleotides upstream of frpA and araC1 genes (Figure 3a) were cloned into the genesgenesirp1 ,irp1irp2, ,irp2irp3, irp3, irp4, irp4, irp5, irp5and andirp9 irp9; outer; outer membrane membrane receptor, receptor,frpA frpA;; transcriptional transcriptional regulators, plasmid pHRP309 upstream of a promoterless lacZ gene. Resulting plasmids were mobilized into V. araC1araC1and andaraC2 araC2; and; and inner inner membrane membrane exporter exporter of of putativeputative siderophore, siderophore, irp8irp8. Predicted. Predicted promoters promoters P1 ordalii andVo-LM-18 P2 containing and the Furtranscription boxes are indicatedlevels of lacZby red were dots measured. RT denotes by determining the location ofβ-galactosidase primer used in P1 and P2 containing Fur boxes are indicated by red dots. RT denotes the location of primer used in activityretrotranscriptase under different reaction conditions while of PCR iron 1, availabilityPCR 2 and PCR(Figure 3 indicate 4). The location use of of the primers PfrpA for (P1) detection and retrotranscriptase reaction while PCR 1, PCR 2 and PCR 3 indicate location of primers for detection ParaC1of (P2) cDNA presumptive from piscibactin promot geneers cluster.produced (b) resultssignificant of three β-galactosidase RT-PCR reactions activity designed when to cells analyze were the of cDNA from piscibactin gene cluster. (b) results of three RT-PCR reactions designed to analyze the culturedtranscription under a strong of the iron irp genelimitation cluster. (CM9 Primer medium marked plus as 2,2’-dipyridylRT, targeted to 80 the µM). 3′-end Under of irp5 iron gene, excess was conditionstranscription (CM9 of or the CM9irp geneplus FeCl cluster.3 25 µM) Primer the marked β-galactosidase as RT, targeted activity toof thethe 3P20-end promoter of irp5 wasgene, 75% was used to obtain a cDNA that spanned from irp5 to araC1. This cDNA was then used as template for ofused the P1 to promoter obtain a cDNA(Figure that 4), spannedsuggesting from a higherirp5 to basalaraC1 activity. This cDNAfor this was promoter. then used However, as template under for three PCR reactions targeted within araC1 (RT-PCR1), frpA (RT-PCR2) and between irp2 3’-end and strongthree iron PCR limitation, reactions targetedthe P2 promoter within araC1 seems(RT-PCR1), to be 10%frpA more(RT-PCR2) active than and P1, between suggestingirp2 3’-enda tighter and irp1 5’-end (RT-PCR3). M, size marker from 100 to 1000 bp. Negative controls (-) are RT-PCR controlirp1 5’-end by iron (RT-PCR3). levels. These M, size results marker demonstrate from 100 to that 1000 the bp. two Negative promoter controls sequences (-) are RT-PCRcould serve reactions as transcriptional lacking reverse starts transcriptase. of the whole Positive irp operon, controls and (that+) are both PCR of reactionsthem are usingstrongly chromosomal regulated by DNA iron as

levelstemplate, with slight+Fe: RT-PCRvariations performed between them. with cells grown under iron excess (TSB-1 + FeCl3 20 µM); -Fe: RT-PCR performed with cells grown under iron limitation (TSB-1 + 2,20-dipyridyl 60 µM).

FigureFigure 4. Transcriptional4. Transcriptional activity activity (β -galactosidase(β-galactosidase units) units) of oflacZ lacZfusions fusions to P1to andP1 and P2 potentialP2 potential promoters of V.ordaliipromoters. β of-galactosidase V ordalii. β-galactosidase activities for activities promoter for P1:: promoterlacZ and P1:: promoterlacZ and P2:: promoterlacZ were P2:: measuredlacZ were in cells culturedmeasured in CM9 in cells minimal cultured medium, in CM9 CM9minimal supplemented medium, CM9 with supplemented 20 µM FeCl with3, as 20 an µM iron FeCl excess3, as condition,an andiron in excess two iron-limiting condition, and conditions: in two iron-limiting CM9 with conditions: 2,20-dipyridyl CM9 with 25 2,2’-dipyridylµM and CM9 25 withµM and 2,2 CM90-dipyridyl 80 µwithM. Three2,2’-dipyridyl independent 80 µM. experiments Three independent were performed experiments in triplicate.were performed Bars represent in triplicate. average Bars values withrepresent standard average deviations valuesindicated with standard by error deviations bars. Theindica datated wereby error analyzed bars. The using data ANOVAwere analyzed signiﬁcance using ANOVA significance test (* p < 0.05). test (* p < 0.05). 3.3. Analysis of Iron-Regulated Outer Membrane Proteins In Gram-negative bacteria some of the outer membrane proteins (OMP) are involved in iron uptake mechanisms, and most of them are regulated by iron. Thus, in order to detect the expression of OMPs involved in siderophore synthesis and transport in V. ordalii we investigated by SDS-PAGE the changes in the OMP profiles when cells were cultured under iron excess or under iron limitation. Some of these proteins could then be identified by PMF. As shown in Figure 5, clear changes in the OMP profile could be detected in three representative strains of V. ordalii when cells were cultured under iron deprivation (the strains were cultured in TSB-1 plus half the MIC of the iron chelator 2,2′-dipyridyl). Five main bands (Table 2) could be identified as proteins clearly regulated by iron

Microorganisms 2019, 7, 313 9 of 16

3.3. Analysis of Iron-Regulated Outer Membrane Proteins In Gram-negative bacteria some of the outer membrane proteins (OMP) are involved in iron uptake mechanisms, and most of them are regulated by iron. Thus, in order to detect the expression of OMPs involved in siderophore synthesis and transport in V. ordalii we investigated by SDS-PAGE the changes in the OMP profiles when cells were cultured under iron excess or under iron limitation. Some of these proteins could then be identified by PMF. As shown in Figure5, clear changes in the OMP profile could be detected in three representative strains of V. ordalii when cells were cultured under iron deprivation (the strains were cultured in TSB-1 plus half the MIC of the iron chelator 2,20-dipyridyl). Microorganisms 2019, 7, x FOR PEER REVIEW 9 of 15 Five main bands (Table2) could be identified as proteins clearly regulated by iron since all them were presentsince all only them in membranewere present fractions only ofin cellsmembrane grown underfractions iron-limiting of cells grown conditions. under Three iron-limiting of these proteinsconditions. were Three high-molecular of these proteins weight were proteins high-mol that wereecular unequivocally weight proteins identified that bywere PMF unequivocally as VabF (311 kDaidentified band markedby PMF asas I VabF in Figure (3115 ),kDa Irp1 band (270 kDamarked band as marked I in Figure as II 5), in FigureIrp1 (2705) and kDa Irp2 band (224 marked kDa band as II markedin Figure as 5) III and in Figure Irp2 5(224). These kDa threeband proteins marked correspond as III in Figure to NRPS 5). These enzymes three involved proteins in vanchrobactincorrespond to (VabF)NRPS andenzymes piscibactin involved (Irp1 in and vanchrobactin Irp2) siderophore (VabF) synthesis. and piscibactin Although NRPSs(Irp1 and are cytosolicIrp2) siderophore enzymes, itsynthesis. has been Although reported thatNRPSs some are of cytosolic them can enzymes, form membrane-bound it has been reported multi-enzymatic that some complexes,of them can called form siderosomes,membrane-bound on the multi-enzymatic inner leaflet of thecomplexes, cytoplasmic called membrane siderosomes, [42,43 on], which the inner could leaflet explain of their the detectioncytoplasmic in membraneV. ordalii membrane [42,43], whic fractions.h could explain Protein their I showed detection 98% in identity V. ordalii to membrane VabF, a NRPS fractions. of V. anguillarumProtein I showedinvolved 98% in vanchrobactin identity to VabF, biosynthesis. a NRPS Proteins of V. IIanguillarum and III clearly involved correspond in vanchrobactin with Irp1 and Irp2,biosynthesis. the two NRPSProteins involved II and III in theclearly synthesis correspond of piscibactin with Irp1 in andP. damselae Irp2, thesubsp. two NRPSpiscicida involved[23,27] in (with the similaritiessynthesis of of piscibactin 70% and 68%, in P. respectively) damselae subsp. and inpiscicidaV. anguillarum [23,27] (with[29] (both similarities proteins of with 70% similarities and 68%, ofrespectively) 98%). and in V. anguillarum [29] (both proteins with similarities of 98%).

Figure 5. RepresentativeRepresentative SDS-PAGE SDS-PAGE gel gel showing showing outer outer membrane membrane protein protein (OMP) (OMP) profiles proﬁles of V. ordalii of V. ordaliistrainsstrains under underiron-rich iron-rich and iron-limited and iron-limited conditions. conditions. MW: Molecular MW: Molecular weight weightmarker; marker; 1: ATCC 1: 33509 ATCCT T T T 33509under iron-excessunder iron-excess conditions; conditions; 2: ATCC 2: 33509 ATCC under 33509 iron-limitationunder iron-limitation (TSB-1 + 2,2 (TSB-1′-dipyridyl+ 2,20-dipyridyl 45 µM); 3: 45Vo-LM-13µM); 3: under Vo-LM-13 iron-excess, under iron-excess, 4: Vo-LM-13 4: Vo-LM-13under iron-limitation under iron-limitation (TSB-1 + 2,2 (TSB-1′-dipyridyl+ 2,2 0-dipyridyl90 µM); 5: 90Vo-LM-18µM); 5: Vo-LM-18under iron-excess; under iron-excess; and 6: Vo-L andM-18 6: Vo-LM-18 under iron-limitation under iron-limitation (TSB-1 + (TSB-12,2′-dipyridyl+ 2,20-dipyridyl 60 µM). 60*: µProteinsM). *: Proteins expressed expressed only unde onlyr underiron ironlimitation limitation and andidentified identiﬁed by byPMF PMF as as follows: follows: I, I, VabF (vanchrobactin(vanchrobactin synthesis);synthesis); II, Irp1; III, Irp2 (pisci (piscibactinbactin synthesis); IV, IV, HuvS (heme receptor); V, FrpA (piscibactin(piscibactin receptor).receptor).

Table 2. Identification by peptide mass fingerprinting (PMF) of five proteins differentially expressed under iron limitation in SDS-PAGE gel showed in Figure 5.

Band in Estimated Similarity Gel (Figure Size Closest Homologues Accession No. (%) 5) (kDa) Band I 311 VabF, V. anguillarum CAJ45639.1 98 Irp1, V. anguillarum WP_019281879.1 98 Band II 270 Irp1, P. damselae subsp. piscicida AKQ52532.1 70 Irp2, V. anguillarum WP_019281878.1 98 Band III 224 Irp2, P. damselae subsp. piscicida AKQ52531.1 68 Band IV 79 HuvS, V. anguillarum CAJ14788.1 99 Band V 71 FrpA, V. anguillarum WP_019281876.1 96

Microorganisms 2019, 7, 313 10 of 16

Table 2. Identification by peptide mass fingerprinting (PMF) of five proteins differentially expressed under iron limitation in SDS-PAGE gel showed in Figure5.

Band in Gel Estimated Size Closest Homologues Accession No. Similarity (%) (Figure5) (kDa) Band I 311 VabF, V. anguillarum CAJ45639.1 98 Irp1, V. anguillarum WP_019281879.1 98 Band II 270 Irp1, P. damselae AKQ52532.1 70 subsp. piscicida Irp2, V. anguillarum WP_019281878.1 98 Band III 224 Irp2, P. damselae AKQ52531.1 68 subsp. piscicida Band IV 79 HuvS, V. anguillarum CAJ14788.1 99 FrpA, V. anguillarum WP_019281876.1 96 Band V 71 FrpA, P. damselae AKQ52529.1 68 subsp. piscicida

The other two diﬀerentially expressed bands with sizes of 79 kDa (band IV in Figure5) and 71 kDa (band V in Figure5) could be identiﬁed as the heme receptor HuvS, showing a 99% similarity to the homologous protein previously reported in V. anguillarum [44], and the piscibactin receptor FrpA, respectively. The later shows a 68% similarity with FrpA protein encoded by the plasmidic irp cluster of P. damselae subsp piscicida [27,40], and a 96% similarity with the FrpA protein reported in V. anguillarum [29]. From the analysis of the iron regulated OMP we could conclude that the irp gene cluster of V. ordalii must be fully functional, since biosynthetic and siderophore transport proteins are detected in cells grown under low iron conditions.

3.4. Identification of Siderophores in Cultures of V. ordalii The genetic and bioinformatic analyses of the irp operon present in V. ordalii, as well as the identification of biosynthetic enzymes encoded by this cluster and induced under iron limitation, strongly indicate that V. ordalii would synthetize the siderophore piscibactin. In order to confirm the synthesis of this siderophore by V. ordalii, cell-free culture supernatants of strain Vo-LM-18 grown under iron-restricted conditions were examined for the presence of piscibactin as described in Material and Methods. The presence of the piscibactin-Ga(III) complex was confirmed on the basis of the accurate mass measurements and the characteristic isotopic cluster of the gallium complex (Figure6). The ∆m/z results for monoisotopic and isotopic ions (M + 1) were below a tolerance acceptable value of 5 ppm and the δ RIA values were within the expected value for positive ion mode (16%) for a compound with a molecular mass range of 350–600 dalton measured in a LTQ-Orbitrap (Table3)[ 45]. From these results, we can unequivocally conclude that piscibactin was present in the culture supernatants of V. ordalii Vo-LM-18. Additionally, vanchrobactin was also detected in the cultures of strain Vo-LM-18 under iron restriction (data not shown). Solid Phase Extraction (SPE) using HLB cartridges of the cell-free culture supernatants of this strain followed by LC-MS analysis showed a peak with a retention time of 4.68 min, + which displays a [M+H] ion at m/z 398.1676 (calculated for C16H24N5O7, m/z 398.1670) in its HRESIMS that corresponds to vanchrobactin [19,29]. Microorganisms 2019, 7, x FOR PEER REVIEW 10 of 15

FrpA, P. damselae subsp. piscicida AKQ52529.1 68

The other two differentially expressed bands with sizes of 79 kDa (band IV in Figure 5) and 71 kDa (band V in Figure 5) could be identified as the heme receptor HuvS, showing a 99% similarity to the homologous protein previously reported in V. anguillarum [44], and the piscibactin receptor FrpA, respectively. The later shows a 68% similarity with FrpA protein encoded by the plasmidic irp cluster of P. damselae subsp piscicida [27,40], and a 96% similarity with the FrpA protein reported in V. anguillarum [29]. From the analysis of the iron regulated OMP we could conclude that the irp gene cluster of V. ordalii must be fully functional, since biosynthetic and siderophore transport proteins are detected in cells grown under low iron conditions.

3.4. Identification of Siderophores in Cultures of V. ordalii The genetic and bioinformatic analyses of the irp operon present in V. ordalii, as well as the identification of biosynthetic enzymes encoded by this cluster and induced under iron limitation, strongly indicate that V. ordalii would synthetize the siderophore piscibactin. In order to confirm the synthesis of this siderophore by V. ordalii, cell-free culture supernatants of strain Vo-LM-18 grown under iron-restricted conditions were examined for the presence of piscibactin as described in Material and Methods. The presence of the piscibactin-Ga(III) complex was confirmed on the basis of the accurate mass measurements and the characteristic isotopic cluster of the gallium complex (Figure 6). The ∆m/z results for monoisotopic and isotopic ions (M + 1) were below a tolerance acceptable value of 5 ppm and the δ RIA values were within the expected value for positive ion mode (16%) for a compound with a molecular mass range of 350–600 dalton measured in a LTQ-Orbitrap (Table 3) [45]. From these results, we can unequivocally conclude that piscibactin was present in the culture supernatants of V. ordalii Vo-LM-18. Microorganisms 2019, 7, 313 11 of 16

RT: 0.00 - 20.00 20160301F9 #1517 RT: 12.04 AV: 1 NL: 6.46E6 10.55 NL: T: FTMS + p ESI Full ms [350.00-600.00] 100 5.66E7 519.99280 10.61 TIC MS z=1 [M+H]+ 90 1.70 2.75 10.53 20160301F 9 6000000 80 2.79 13.26 5500000 2.80 521.99127 70 10.64 5000000 z=1 (a) 10.66 13.89 4500000 (c) 60 4000000 2.86 50 10.68 3500000 2.89 15.88 16.43 1.19 12.06 7.44 7.47 3000000 40 7.49 Intensity 4.38 2500000 7.41 7.54 30 4.34 4.43 15.28 7.36 7.58 16.94 2000000 19.82 520.99603 4.48 7.33 7.66 20 9.83 1500000 z=1 6.24 522.99481 9.62 z=1 5.41 1000000 517.00214 10 z=1 500000 518.00494 518.99719 z=? z=? 0 0 0 2 4 6 8 10 12 14 16 18 20 516 517 518 519 520 521 522 523 Time (min) m/z RT: 0.00 - 20.00 20160301F9#1525 RT: 12.11 12.06 NL: 100 7.15E6 T: FTMS + p ESI Full ms [350.00-600.00] 12.08 m/z= m/z= 516.0000-523.5000 90 12.02 519.70- m/z Intensity Relative Charge 520.40 MS 20160301F 516.0084 76070.0 1.30 1.00 80 12.14 9 517.0021 233827.2 3.99 1.00 70 517.2151 55866.4 0.95 0.00 (b) 12.16 60 518.0051 47397.8 0.81 0.00 11.96 12.19 519.9928 5863022.5 100.00 1.00 (d) 50 520.0153 25896.3 0.44 0.00 520.9960 1138102.0 19.41 1.00 40 521.9913 4154611.3 70.86 1.00 12.23 30 522.9944 803882.6 13.71 1.00

12.32 10 15.86 12.60 1.70 2.46 5.37 6.36 9.71 11.84 15.97 18.62 0 0 2 4 6 8 10 12 14 16 18 20 Time (min)

Figure 6. (a) LC-HRMS total ion chromatogram of the fraction containing piscibactin-Ga(III) complex ® eluted with H2O and CH3CN (1:1) from an OASIS HLB cartridge; (b) extracted ion chromatogram from m/z 519.7–520.4; (c) high resolution mass spectrum corresponding to the peak with t = 12.06 R 13 12 min; (d) isotopic ion abundance from m/z 516.0–523.5, including M+1/M( C1/ C), calculated with Thermo Xcalibur software 3.0. LTQ-Orbitrap was operating at a resolving power of 30,000 (m/∆m) with a detection window setting around the compound of interest (between 350 and 600 dalton).

a Table3. List of ions observed in fraction eluted with H2O and CH3CN (1:1) , using LC-ESI-LTQ-Orbitrap in Positive-ion Mode b.

Retention Detected Mean Ion ∆ m/z (ppm) Ion Formula δRIA (%) Time (min) [M+H]+ Intensity 519.99280 3.1 12C H 69GaN O S + 5.8 106 - 19 21 3 4 3 × 520.99603 3.4 13C 12C H 69GaN O S + 1.1 106 5.3 Piscibactin-Ga(III) 12.08 1 18 21 3 4 3 × − 521.99127 4.4 12C H 71GaN O S + 4.6 106 - 19 21 3 4 3 × 522.99481 4.1 13C 12C H 71GaN O S + 8.0 105 5.9 1 18 21 3 4 3 × a b Fraction mass 16.9 mg. Fraction L3 was mixed with 3 µL of a 12 mg/mL solution of GaBr3/H2O just before injection.

4. Discussion V. ordalii is the causative agent of vibriosis in several salmonid fish species farmed in several geographic areas around the world [1]. Although it was formerly classified as V. anguillarum biovar II, it was later recognized as a new Vibrio species [46]. Despite the similarities between both species, each of them causes quite different types of vibriosis [1,47] and some important genomic differences were reported between both species, for example the size of the genome of V. ordalii being 70% of that of V. anguillarum [4]. Both species also present important phenotypic differences [1]. Thus, the differences could also reach the virulence mechanisms used by each one to cause disease in fish. Among the variety of virulence factors present in V. anguillarum, the iron uptake systems are among the best studied [1,48]. However, these mechanisms are yet poorly known in V. ordalii. In a previous work we could detect the production of siderophores and suggested that piscibactin could be a siderophore Microorganisms 2019, 7, 313 12 of 16 being produced by this bacterium [26]. In the present research we could demonstrate that piscibactin is really the siderophore synthesized by V. ordalii under iron deprivation. Vanchrobactin is a chromosomally encoded siderophore that is conserved among all V. anguillarum isolates as either environmental or pathogenic [21,49]. As noted above, all the genes necessary for vanchrobactin synthesis are also present in the genome of V. ordalii [4] and VabF, the NRPS that ensembles vanchrobactin, can be actually detected under low iron conditions (Figure5). Moreover, we could detect the presence of vanchrobactin in the supernatants of V.ordalii cultured under iron limitation (data not shown). However, part of the required transporters, specifically the ABC transporters fvtB-fvtE, seem to be missed from the genome of V. ordalii [4]. Furthermore, although fvtA, the gene encoding the vanchrobactin outer membrane receptor of V. anguillarum [49], is present in the genome of V. ordalii, we could never detect any homolog to FvtA in the V. ordalii OMP profiles, suggesting that fvtA is not expressed. Thus, although V. ordalii also produces vanchrobactin, our results suggest that this bacterium is unable to use it as siderophore, confirming previously reported genomic and biological studies [4,26]. In addition to vanchrobactin, some strains of V. anguillarum lacking pJM1-type plasmids (that encode the synthesis of anguibactin [50]) produce also piscibactin as siderophore. The synthesis of piscibactin in V. anguillarum is favored at low temperatures since the transcriptional activity of the biosynthetic genes is three-times higher at 18 ◦C than at 25 ◦C[29]. Although in V. anguillarum vanchrobactin and piscibactin are simultaneously produced, the latter is a key virulence factor to infect fish whereas vanchrobactin seems to have a secondary role in virulence. This is in agreement with the observation that piscibactin seems to be the only siderophore used for iron uptake by V. ordalii. The fact that piscibactin synthesis in V. anguillarum is preferentially expressed below 18 ◦C also agrees with the usually lower optimal growth temperature of V. ordalii compared to V. anguillarum [2,46,51]. Thus, synthesis of piscibactin could be an adaptation to infect hosts that grow at low temperatures, and in these conditions piscibactin could be an efficient siderophore. It is noteworthy that piscibactin is the siderophore present in more species within the Vibrionaceae family than any other siderophore system. This wide distribution of piscibactin could be explained by a horizontal gene transfer (HGT) event that was followed by the action of diverse evolutionary forces [28]. As demonstrated in P. damselae subsp. piscicida, piscibactin is encoded by a pathogenicity island which resembles the high pathogenicity island (HPI) encoding the siderophore yersiniabactin in Yersinia [27,40]. The plasmid harboring this pathogenicity island could be transferred to other marine bacteria [27]. Acquisition of piscibactin genes by HGT could lead to the inactivation of other siderophore systems present in the ancient Vibrio genome. A similar event was demonstrated in V. anguillarum, in which the acquisition of the pJM1 plasmid, encoding the anguibactin siderophore system, led to the inactivation of vanchrobactin synthesis by a transposon harbored by the plasmid [21,52]. It is likely that the siderophore with the highest affinity for iron could have a selective advantage. V. ordalii contains a significantly smaller genome than V. anguillarum, which explains the physical and ecological differences existing between both species [4]. Besides, this reduced genome suggests that V. ordalii may be immersed in the process of evolution toward an endosymbiotic lifestyle [5]. In this scenario, it is likely that vanchrobactin synthesis does not have any advantage, since its production could be more related to persistence into a marine environment than to pathogenesis [21,29]. Since piscibactin is a key virulence factor for V. anguillarum strains lacking the anguibactin system, and due to the close genetic relationship between both species, it is reasonable to speculate that the same will be true for V. ordalii. Multiple attempts to generate V. ordalii knock-out mutants (by the allelic exchange method previously used for V. anguillarum [29]) defective in piscibactin production were unsuccessful (data not shown). Further research is needed to try to generate piscibactin-deficient mutants in this bacterium to clearly demonstrate the involvement of piscibactin in the pathogenesis of vibriosis caused by V. ordalii. In conclusion, V. ordalii produces piscibactin and vanchrobactin as siderophores when it is cultured under low iron conditions, but only piscibactin is used for iron uptake. The fact that piscibactin is a key Microorganisms 2019, 7, 313 13 of 16 virulence factor in other fish pathogens like P. damselae subsp. piscicida and V. anguillarum, highlights the importance of this siderophore in the pathogenesis of diseases caused by Vibrionaceae members in poikilothermic animals.

Author Contributions: Conceptualization, M.B., J.R., C.J., R.A.-H. and M.L.L.; data curation, M.B. and J.C.F.-M.; funding acquisition, J.R., C.J., R.A.-H. and M.L.L.; investigation, P.R., M.B. and J.C.F.-M.; methodology, P.R., M.B. and J.C.F.-M.; project administration, A.E.T.; supervision, A.E.T., J.R., C.J., R.A.-H. and M.L.L.; validation, A.E.T.; visualization, P.R., M.B., J.C.F.-M. and J.R.; writing—original draft, P.R., R.A.-H. and M.L.L.; writing—review and editing, C.J., R.A.-H. and M.L.L. Funding: This work was supported by Grants CONICYT/FONDAP/15110027 and FONDECYT Nº 1190283 from the Comisión Nacional de Investigación Científica y Tecnológica (CONICYT, Chile), and by grants AGL2015-63740-C2-1-R and AGL2015-63740-C2-2-R from the State Agency for Research (AEI) of Spain, both co-funded by the FEDER Programme from the European Union. Grants from Xunta de Galicia (Spain) supporting work at University of Santiago de Compostela (grant GRC2018/18) and University of A Coruña (grant GRC2018/39 and CICA-INIBIC support ED431E2018/03) are also acknowledged. Acknowledgments: Pamela Ruiz acknowledges reception of a CONICYT Doctoral Scholarship (No. 21110146). Ruben Avendaño-Herrera acknowledges Rodolfo Paredes (from Universidad Andrés Bello, Chile) for his help in protein analysis. Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

References

1. Toranzo, A.E.; Magariños, B.; Avendaño-Herrera, R. Vibriosis: Vibrio anguillarum, V. ordalii and Aliivibrio salmonicida. In Fish Viruses and Bacteria: Pathobiology and Protection; Woo, P.T.K., Cipriano, R.C., Eds.; CABI: London, UK, 2017; pp. 314–333. 2. Colquhoun, D.J.; Aase, I.L.; Wallace, C.; Baklien, A.; Gravningen, K. First description of Vibrio ordalii from Chile. Bull. Eur. Assoc. Fish. Pathol. 2004, 24, 185–188. 3. Ruiz, P.; Poblete, M.; Yañez, A.J.; Irgang, R.; Toranzo, A.E.; Avendaño-Herrera, R. Cell-surface properties of Vibrio ordalii strains isolated from Atlantic salmon Salmo salar in Chilean farms. Dis. Aquat. Organ. 2015, 113, 9–23. [CrossRef][PubMed] 4. Naka, H.; Dias, G.M.; Thompson, C.C.; Dubay, C.; Thompson, F.L.; Crosa, J.H. Complete genome sequence of the marine fish pathogen Vibrio anguillarum harboring the pJM1 virulence plasmid and genomic comparison with other virulent strains of V. anguillarum and V. ordalii. Infect. Immun. 2011, 79, 2889–2900. [CrossRef] [PubMed] 5. Steinum, T.M.; Karatas, S.; Martinussen, N.T.; Meirelles, P.M.; Thompson, F.L.; Colquhoun, D.J. Multilocus Sequence Analysis of Close Relatives Vibrio anguillarum and Vibrio ordalii. Appl. Environ. Microbiol. 2016, 82, 5496–5504. [CrossRef][PubMed] 6. Ruiz, P.; Poblete-Morales, M.; Irgang, R.; Toranzo, A.E.; Avendaño-Herrera, R. Survival behaviour and virulence of the fish pathogen Vibrio ordalii in seawater microcosms. Dis. Aquat. Organ. 2016, 120, 27–38. [CrossRef][PubMed] 7. Miethke, M. Molecular strategies of microbial iron assimilation: From high-affinity complexes to cofactor assembly systems. Metallomics 2013, 5, 15–28. [CrossRef][PubMed] 8. Skaar, E.P.; Raffatellu, M. Metals in infectious diseases and nutritional immunity. Metallomics 2015, 7, 926–928. [CrossRef][PubMed] 9. Wilson, B.R.; Bogdan, A.R.; Miyazawa, M.; Hashimoto, K.; Tsuji, Y. Siderophores in Iron Metabolism: From Mechanism to Therapy Potential. Trends Mol. Med. 2016, 22, 1077–1090. [CrossRef][PubMed] 10. Hood, M.I.; Skaar, E.P. Nutritional immunity: Transition metals at the pathogen-host interface. Nat. Rev. Microbiol. 2012, 10, 525–537. [CrossRef] 11. Fillat, M.F. The FUR (ferric uptake regulator) superfamily: Diversity and versatility of key transcriptional regulators. Arch. Biochem. Biophys. 2014, 546, 41–52. [CrossRef] 12. Richard, K.L.; Kelley, B.R.; Johnson, J.G. Heme Uptake and Utilization by Gram-Negative Bacterial Pathogens. Front. Cell. Infect. Microbiol. 2019, 9, 81. [CrossRef][PubMed] Microorganisms 2019, 7, 313 14 of 16

13. Holden, V.I.; Bachman, M.A. Diverging roles of bacterial siderophores during infection. Metallomics 2015, 7, 986–995. [CrossRef][PubMed] 14. Ellermann, M.; Arthur, J.C. Siderophore-mediated iron acquisition and modulation of host-bacterial interactions. Free Radic. Biol. Med. 2017, 105, 68–78. [CrossRef][PubMed] 15. Stork, M.; Di Lorenzo, M.; Mouriño, S.; Osorio, C.R.; Lemos, M.L.; Crosa, J.H. Two tonB systems function in iron transport in Vibrio anguillarum, but only one is essential for virulence. Infect. Immun. 2004, 72, 7326–7329. [CrossRef][PubMed] 16. Kuehl, C.J.; Crosa, J.H. The TonB energy transduction systems in Vibrio species. Future Microbiol. 2010, 5, 1403–1412. [CrossRef][PubMed] 17. Krewulak, K.D.; Vogel, H.J. TonB or not TonB: Is that the question? Biochem. Cell Biol. 2011, 89, 87–97. [CrossRef][PubMed] 18. Avendaño-Herrera, R.; Toranzo, A.E.; Romalde, J.L.; Lemos, M.L.; Magariños, B. Iron uptake mechanisms in the fish pathogen Tenacibaculum maritimum. Appl. Environ. Microbiol. 2005, 71, 6947–6953. [CrossRef] [PubMed] 19. Soengas, R.G.; Anta, C.; Espada, A.; Paz, V.; Ares, I.R.; Balado, M.; Rodríguez, J.; Lemos, M.L.; Jiménez, C. Structural characterization of vanchrobactin, a new catechol siderophore produced by the fish pathogen Vibrio anguillarum serotype O2. Tetrahedron Lett. 2006, 47, 7113–7116. [CrossRef] 20. Lemos, M.L.; Osorio, C.R. Heme, an iron supply for vibrios pathogenic for fish. Biometals 2007, 20, 615–626. [CrossRef] 21. Lemos, M.L.; Balado, M.; Osorio, C.R. Anguibactin- versus vanchrobactin-mediated iron uptake in Vibrio anguillarum: Evolution and ecology of a fish pathogen. Environ. Microbiol. Rep. 2010, 2, 19–26. [CrossRef] 22. Retamales, J.; González-Contreras, A.; Salazar, S.; Toranzo, A.E.; Avendaño-Herrera, R. Iron utilization and siderophore production by Streptococcus phocae isolated from diseased Atlantic salmon (Salmo salar). Aquaculture 2012, 364–365, 305–311. [CrossRef] 23. Souto, A.; Montaos, M.A.; Rivas, A.J.; Balado, M.; Osorio, C.R.; Rodríguez, J.; Lemos, M.L.; Jiménez, C. Structure and Biosynthetic Assembly of Piscibactin, a Siderophore from Photobacterium damselae subsp. piscicida, Predicted from Genome Analysis. Eur. J. Org. Chem. 2012, 2012, 5693–5700. 24. Balado, M.; Souto, A.; Vences, A.; Careaga, V.P.; Valderrama, K.; Segade, Y.; Rodríguez, J.; Osorio, C.R.; Jiménez, C.; Lemos, M.L. Two Catechol Siderophores, Acinetobactin and Amonabactin, Are Simultaneously Produced by Aeromonas salmonicida subsp. salmonicida Sharing Part of the Biosynthetic Pathway. ACS Chem. Biol. 2015, 10, 2850–2860. [PubMed] 25. Bethke, J.; Poblete-Morales, M.; Irgang, R.; Yañez, A.; Avendaño-Herrera, R. Iron acquisition and siderophore production in the fish pathogen Renibacterium salmoninarum. J. Fish. Dis. 2016, 39, 1275–1283. [CrossRef] [PubMed] 26. Ruiz, P.; Balado, M.; Toranzo, A.E.; Poblete-Morales, M.; Lemos, M.L.; Avendaño-Herrera, R. Iron assimilation and siderophore production by Vibrio ordalii strains isolated from diseased Atlantic salmon Salmo salar in Chile. Dis. Aquat. Organ. 2016, 118, 217–226. [CrossRef] 27. Osorio, C.R.; Rivas, A.J.; Balado, M.; Fuentes-Monteverde, J.C.; Rodríguez, J.; Jiménez, C.; Lemos, M.L.; Waldor, M.K. A Transmissible Plasmid-Borne Pathogenicity Island Confers Piscibactin Biosynthesis in the Fish Pathogen Photobacterium damselae subsp. piscicida. Appl. Environ. Microbiol. 2015, 81, 5867–5879. [CrossRef][PubMed] 28. Thode, S.K.; Rojek, E.; Kozlowski, M.; Ahmad, R.; Haugen, P. Distribution of siderophore gene systems on a Vibrionaceae phylogeny: Database searches, phylogenetic analyses and evolutionary perspectives. PLoS ONE 2018, 13, e0191860. [CrossRef] 29. Balado, M.; Lages, M.A.; Fuentes-Monteverde, J.C.; Martínez-Matamoros, D.; Rodríguez, J.; Jiménez, C.; Lemos, M.L. The Siderophore Piscibactin Is a Relevant Virulence Factor for Vibrio anguillarum Favored at Low Temperatures. Front. Microbiol. 2018, 9, 1766. [CrossRef][PubMed] 30. Avendaño-Herrera, R.; Maldonado, J.P.; Tapia-Cammas, D.; Feijoo, C.G.; Calleja, F.; Toranzo, A.E. PCR protocol for detection of Vibrio ordalii by amplification of the vohB (hemolysin) gene. Dis. Aquat. Organ. 2014, 107, 223–234. [CrossRef] Microorganisms 2019, 7, 313 15 of 16

31. Lemos, M.L.; Salinas, P.; Toranzo, A.E.; Barja, J.L.; Crosa, J.H. Chromosome-mediated iron uptake system in pathogenic strains of Vibrio anguillarum. J. Bacteriol. 1988, 170, 1920–1925. [CrossRef] 32. Solovyev, V.; Salamov, A. Automatic Annotation of Microbial Genomes and Metagenomic Sequences. In Metagenomics and Its Applications in Agriculture, Biomedicine and Environmental Studies; Li, R.W., Ed.; Nova Science Publishers: New York, NY, USA, 2011; pp. 61–78. 33. Escolar, L.; Perez-Martin, J.; de Lorenzo, V. Opening the iron box: Transcriptional metalloregulation by the Fur protein. J. Bacteriol. 1999, 181, 6223–6229. [PubMed] 34. Parales, R.E.; Harwood, C.S. Construction and use of a new broad-host-range lacZ transcriptional fusion vector, pHRP309, for Gram- bacteria. Gene 1993, 133, 23–30. [CrossRef] 35. Miller, J.H. Experiments in Molecular Genetics; Cold Spring Harbor Laboratory: Cold Spring Harbor, NY, USA, 1972. 36. Crosa, J.H.; Hodges, L.L. Outer membrane proteins induced under conditions of iron limitation in the marine fish pathogen Vibrio anguillarum 775. Infect. Immun. 1981, 31, 223–227. [PubMed] 37. Toranzo, A.E.; Barja, J.L.; Potter, S.A.; Colwell, R.R.; Hetrick, F.M.; Crosa, J.H. Molecular factors associated with virulence of marine vibrios isolated from striped bass in Chesapeake Bay. Infect. Immun. 1983, 39, 1220–1227. [PubMed] 38. Puentes, B.; Balado, M.; Bermudez-Crespo, J.; Osorio, C.R.; Lemos, M.L. A proteomic analysis of the iron response of Photobacterium damselae subsp. damselae reveals metabolic adaptations to iron levels changes and novel potential virulence factors. Vet. Microbiol. 2017, 201, 257–264. 39. Finn, R.D.; Bateman, A.; Clements, J.; Coggill, P.; Eberhardt, R.Y.; Eddy, S.R.; Heger, A.; Hetherington, K.; Holm, L.; Mistry, J.; et al. Pfam: The protein families database. Nucleic Acids Res. 2014, 42, D222–D230. [CrossRef][PubMed] 40. Osorio, C.R.; Juiz-Rio, S.; Lemos, M.L. A siderophore biosynthesis gene cluster from the fish pathogen Photobacterium damselae subsp. piscicida is structurally and functionally related to the Yersinia high-pathogenicity island. Microbiology 2006, 152, 3327–3341. 41. Crosa, J.H.; Walsh, C.T. Genetics and assembly line enzymology of siderophore biosynthesis in bacteria. Microbiol. Mol. Biol. Rev. 2002, 66, 223–249. [CrossRef] 42. Imperi, F.; Visca, P. Subcellular localization of the pyoverdine biogenesis machinery of Pseudomonas aeruginosa: A membrane-associated “siderosome”. FEBS Lett. 2013, 587, 3387–3391. [CrossRef] 43. Gasser, V.; Guillon, L.; Cunrath, O.; Schalk, I.J. Cellular organization of siderophore biosynthesis in Pseudomonas aeruginosa: Evidence for siderosomes. J. Inorg. Biochem. 2015, 148, 27–34. [CrossRef] 44. Mouriño, S.; Rodríguez-Ares, I.; Osorio, C.R.; Lemos, M.L. Genetic variability of the heme uptake system among different strains of the fish pathogen Vibrio anguillarum: Identification of a new heme receptor. Appl. Environ. Microbiol. 2005, 71, 8434–8441. [CrossRef][PubMed] 45. Xu, Y.; Heilier, J.F.; Madalinski, G.; Genin, E.; Ezan, E.; Tabet, J.C.; Junot, C. Evaluation of accurate mass and relative isotopic abundance measurements in the LTQ-orbitrap mass spectrometer for further metabolomics database building. Anal. Chem. 2010, 82, 5490–5501. [CrossRef][PubMed] 46. Schiewe, M.H.; Trust, T.J.; Crosa, J.H. Vibrio ordalii sp. nov.: A causative agent of vibriosis in fish. Curr. Microbiol. 1981, 6, 343–348. [CrossRef] 47. Ransom, D.P.; Lannan, C.N.; Rohovec, J.S.; Fryer, J.L. Comparison of histopathology caused by Vibrio anguillarum and Vibrio ordalii in three species of Pacific salmon. J. Fish. Dis. 1984, 7, 107–115. [CrossRef] 48. Li, Y.; Ma, Q. Iron Acquisition Strategies of Vibrio anguillarum. Front. Cell. Infect. Microbiol. 2017, 7, 342. [CrossRef][PubMed] 49. Balado, M.; Osorio, C.R.; Lemos, M.L. FvtA is the receptor for the siderophore vanchrobactin in Vibrio anguillarum: Utility as a route of entry for vanchrobactin analogues. Appl. Environ. Microbiol. 2009, 75, 2775–2783. [CrossRef][PubMed] 50. Stork, M.; Di Lorenzo, M.; Welch, T.J.; Crosa, L.M.; Crosa, J.H. Plasmid-mediated iron uptake and virulence in Vibrio anguillarum. Plasmid 2002, 48, 222–228. [CrossRef] Microorganisms 2019, 7, 313 16 of 16

51. Actis, L.A.; Tolmasky, M.E.; Crosa, J.H. Vibriosis. In Fish Diseases and Disorders; Woo, P.T.K., Bruno, D.W., Eds.; CAB International: London, UK, 2011; Volume 3, pp. 570–604. 52. Naka, H.; Lopez, C.S.; Crosa, J.H. Reactivation of the vanchrobactin siderophore system of Vibrio anguillarum by removal of a chromosomal insertion sequence originated in plasmid pJM1 encoding the anguibactin siderophore system. Environ. Microbiol. 2008, 10, 265–277. [CrossRef][PubMed]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).